Most pharmaceutical solutions and suspensions manufactured on an industrial scale require highly controlled, thorough mixing to achieve a satisfactory yield and to ensure a uniform distribution of ingredients in the final product. Agitator tanks are frequently used to complete the mixing process, but a better degree of mixing is normally achieved by using a mechanical stirrer or impeller (e.g., a set of mixing blades attached to a metal rod). Typically, the mechanical stirrer or impeller is simply lowered into the fluid through an opening in the top of the vessel and rotated by an external motor to create the desired mixing action.
One significant limitation or shortcoming of such an arrangement is the danger of contamination or leakage during mixing. The rod carrying the mixing blades or impeller is typically introduced into the vessel through a dynamic seal or bearing. This opening provides an opportunity for bacteria or other contaminants to enter, which of course can lead to the degradation of the product. A corresponding danger of environmental contamination exists in applications involving hazardous or toxic fluids, or suspensions of pathogenic organisms, since dynamic seals or bearings are prone to leakage. Cleanup and sterilization are also made difficult by the dynamic bearings or seals, since these structures typically include folds and crevices that are difficult to reach. Since these problems are faced by all manufacturers of sterile solutions, pharmaceuticals, or the like, the U.S. Food and Drug Administration (FDA) has consequently promulgated strict processing requirements for such fluids, and especially those slated for intravenous use.
Recently, there has also been an extraordinary increase in the use of biosynthetic pathways in the production of pharmaceutical materials, but problems plague those involved in this rapidly advancing industry. The primary problem is that suspensions of genetically altered bacterial cells frequently used to produce protein pharmaceuticals (insulin is a well-known example) require gentle mixing to circulate nutrients. If overly vigorous mixing or contact between the impeller and the vessel wall occurs, the resultant forces and shear stresses may damage or destroy a significant fraction of the cells, as well as protein molecules that are sensitive to shear stresses. This not only reduces the beneficial yield of the process, but also creates deleterious debris in the fluid suspension that requires further processing to remove.
In an effort to overcome this problem, others have proposed alternative mixing technologies. The most common proposal for stirring fluids under sterile conditions is to use a rotating, permanent magnet bar covered by an inert layer of TEFLON, glass, or the like. The magnetic bar is placed on the bottom of the agitator vessel and rotated by a driving magnet positioned external to the vessel. Of course, the use of such an externally driven magnetic bar avoids the need for a dynamic bearing, seal or other opening in the vessel to transfer the rotational force from the driving magnet to the stirring magnet. Therefore, a completely enclosed system is provided. This of course prevents leakage and the potential for contamination created by hazardous materials (e.g., cytotoxic agents, solvents with low flash points, blood products, etc.), eases clean up, and allows for the desirable sterile interior environment to be maintained.
However, several well-recognized drawbacks are associated with this mixing technology, making it unacceptable for use in many applications. For example, the driving magnet produces not only torque on the stirring magnetic bar, but also an attractive axial thrust force tending to drive the bar into contact with the bottom wall of the vessel. This of course generates substantial friction at the interface between the bar and the bottom wall of the vessel. This uncontrolled friction generates unwanted heat and may also introduce an undesirable shear stress in the fluid. Consequently, fragile biological molecules, such as proteins and living cells that are highly sensitive to temperature and shear stress, are easily damaged during the mixing process, and the resultant debris may contaminate the product. Moreover, the magnetic bar stirrer may not generate the level of circulation provided by an impeller, and thus cannot be scaled up to provide effective mixing throughout the entire volume of large agitation tanks of the type preferred in commercial production operations.
In yet another effort to eliminate the need for dynamic bearings or shaft seals, some have proposed mixing vessels having external magnets that remotely couple the mixing impeller to a motor located externally to the vessel. A typical magnetic coupler comprises a drive magnet attached to the motor and a stirring magnet carrying an impeller. Similar to the magnetic bar technology described above, the driver and stirrer magnets are kept in close proximity to ensure that the coupling between the two is strong enough to provide sufficient torque. An example of one such proposal is found in U.S. Pat. No. 5,470,152 to Rains.
As described above, the high torque generated can drive the impeller into the walls of the vessel creating significant friction. By strategically positioning roller bearings inside the vessel, the effects of friction between the impeller and the vessel wall can be substantially reduced. Of course, high stresses at the interfaces between the ball bearings and the vessel wall or impeller result in a grinding of the mixing proteins and living cells, and loss of yield. Further, the bearings may be sensitive to corrosive reactions with water-based solutions and other media and will eventually deteriorate, resulting in frictional losses that slow the impeller, reduce the mixing action, and eventually also lead to undesirable contamination of the product. Bearings also add to the cleanup problems.
In an effort to address and overcome the limitations described above, still others have proposed levitated bearings designed to reduce the deleterious effects of friction resulting from magnetically coupled mixers. By using a specially configured magnetic coupler to maintain only a repulsive levitation force in the vertical direction, the large thrust force between the stirring and driving magnets can be eliminated, along with the resultant shear stress and frictional heating. An example of one such arrangement is shown in U.S. Pat. No. 5,478,149 to Quigg.
However, one limitation remaining from this approach is that only magnet-magnet interactions provide the levitation. This leads to intrinsically unstable systems that produce the desired levitation in the vertical direction, but are unable to control side-to-side movement. As a result, external contact bearings in the form of bearing rings are necessary to laterally stabilize the impeller. Although this “partial” levitation reduces the friction between the impeller and the vessel walls, it does not totally eliminate the drawbacks of the magnetically coupled, roller bearing mixers previously mentioned.
In an effort to eliminate the need for contact or other types of mechanical roller bearings, complex feedback control has been proposed to stabilize the impeller. Typical arrangements use electromagnets positioned alongside the levitating magnet. However, the high power level required to attain only sub-millimeter separations between the levitating magnet and the stabilizing magnets constitutes a major disadvantage of this approach. Furthermore, this solution is quite complex, since the stabilizing magnets must be actively monitored and precisely controlled by complex computer-implemented software routines to achieve even a moderate degree of stability. As a consequence of this complexity and the associated maintenance expense, this ostensible solution has not been accepted in the commercial arena, and it is doubtful that it can be successfully scaled up for use in mixing industrial or commercial scale process volumes.
Still others have proposed the use of superconductive materials to levitate magnetic bearings. Despite recent advances in the art, significant limitation on the application of this technology to mixing systems results from the extraordinarily cold temperatures required to create the desired superconductive effects. Even the recently discovered “high temperature” superconductors require temperatures on the order of 77 to 130 Kelvin to induce reliable, stable levitation in a magnetic bearing. In the past, the relatively wide separation distance required between the bearing, the cryostat outer wall, and the superconducting element necessary to prevent unwanted cooling of the fluid has limited the industrial applicability of this approach. To date, applications of this technology to fluids have been primarily in the pumping of cryogens or the like, such as those typically used in cold fusion experiments, in flywheels or other energy storage devices, or for space travel (see representative U.S. Pat. No. 5,747,426 to Abboud or U.S. Pat. No. 4,365,942 to Schmidt), where there is of course little concern for the inevitable cooling effect created.
In my prior U.S. Pat. No. 5,567,672, I describe levitating a magnet above a superconducting element in a cryostat, which contains the cooling source used to cool the superconducting element. This arrangement could possibly be used as part of a system for mixing temperature sensitive fluids, such as cell suspensions or blood, as disclosed herein. However, the resultant increased separation created by the double wall vacuum gap may decrease the stability and the load capacity of the levitating magnet. This may limit the applications in which this arrangement is useful, and could especially preclude use with particularly viscous fluids or with the large volumes of fluid typically present in commercial scale operations.
Thus, a need is identified for an improved system having a levitating magnetic bearing for mixing or pumping fluids, and especially ultra-pure, hazardous, or delicate fluid solutions or suspensions. The system would preferably employ a magnetic bearing that levitates in a stable fashion to avoid contact with the bottom or side walls of the vessel. Since the bearing would levitate in the fluid, no mixing rod or other structure penetrating the mixing vessel would be necessary, thus eliminating the need for dynamic bearings or seals and all potentially deleterious effects associated therewith. Since penetration is unnecessary, the vessel could be completely sealed prior to mixing to avoid the potential for contamination and reduce the chance for exposure in the case of hazardous or biological fluids, such as contaminated blood or the like. The vessel and magnetic bearing could also be made of disposable materials and discarded after each use, which would eliminate the need for cleaning or sterilization. The absence of a mixing or stirring rod penetrating through the vessel would also allow a slowly rotating impeller to be held at an off-axis position in a sealed vessel, thus making it possible to independently rotate the vessel about its central axis to achieve very gentle, yet thorough, mixing.
In the case of warm or temperature-sensitive fluids, the use of superconductivity to provide the desired levitation would be possible by thermally isolating and separating the superconducting element from the magnetic bearing and providing a separate, substantially isolated cooling source. This combined thermal isolation and separation would avoid creating any significant cooling in the vessel, the magnetic bearing or the fluid being mixed or pumped. The use of a superconductor would also eliminate the sole reliance on magnet-magnet repulsion to provide the levitation force and the concomitant need for active electronic control systems to ensure stable levitation. Overall, the proposed system would have superior characteristics over existing mixing or pumping technologies, especially in terms of sterility, mixing quality, safety and reliability, and would be readily adaptable for use in larger, industrial scale operations.